Podcast Summary: How I AI – Guillermo Rauch: Vercel CEO on how v0 hit 3,200 PRs merged per day (and lets anyone ship)

Host: Claire Vo
Guest: Guillermo Rauch (CEO, Vercel)
Date: February 4, 2026
Episode Theme:
How Vercel’s new v0 platform is revolutionizing software development—democratizing the ability for anyone, technical or not, to ship production-grade changes fast. Guillermo dives into practical AI-driven workflows, internal best practices, and real-world examples of v0's power, scale, and future applications.

1. Main Theme & Purpose

• The episode focuses on actionable, real-world use of AI in software development, showcasing how Vercel's v0 tool dramatically lowers the barrier from idea to production, letting anyone—from marketers to engineers—ship changes safely and quickly.
• Guillermo shares Vercel's internal practices, how they foster a culture where anyone can "cook" (experiment and build) with v0, the surge in usage (3,200+ PRs merged per day), and the ripple effects on speed and collaboration across organizations.

2. Key Discussion Points & Insights

The Old vs. New Paradigm of Shipping Changes

• High Friction in the Past:
  Shipping even simple changes (like adding a logo) required petitioning engineers or praying the CMS was flexible enough.
  Quote: "One is what I call you had a petition to the government, you had to go to engineers and say, engineers, can you please add a logo over here or whatnot."
  (Guillermo, 00:15)

• v0’s Impact:
  Anyone can now open a page in v0 and “prompt” what they want, slashing friction and letting ideas move straight from thought to production, with safeguards. Quote: "Now they can just open this page in V0 and prompt anything that they want."
  (Guillermo, 00:34)

• Cultural Shift:
  Prioritization rituals (red tape) disappear, allowing teams to focus on defending ideas by their merits. Quote: "The humiliation ritual of prioritization goes away and you can actually focus your time on defending the merits of an idea on the actual idea..."
  (Claire, 00:38, 20:37)

Skills.sh – v0’s Viral AI Skills Ecosystem

• Creation and Ecosystem:
  Vercel launched Skills.sh, an open hub akin to "npm for AI skills," enabling rapid adoption and sharing of agentic workflows, best practices, and modular capabilities. Quote: "Skills Sh is a new… You can think of it as like npm. It's a hub, an open ecosystem of skills."
  (Guillermo, 01:41)

• Scale and Community Adoption:
  34,000+ skills submitted (and growing by 500/hour), with usage clustering at the top but strong community engagement. Quote: "It's pretty dramatic what's happening to this site. So you can see that we now have 34,000 skills submitted by the community and this website has gone viral all over the Internet."
  (Guillermo, 01:51)

• Personal Uses:
  Claire uses community skills to rapidly build video components and automate processes she otherwise couldn't. Quote: "Let me import components and regularly create videos really, really quickly. And I would not have been able to do this without the expertise that’s been packaged in these best practices..."
  (Claire, 02:54)

Practical Workflow: Iterating & Shipping with v0

• Branch-First Development:
  v0 integrates a Git-like flow—branches, previews, code review, and PRs—right into the UI, making production-grade engineering dead simple for anyone. Quote: "V0 is making the git flow of creating branches a first class citizen of the product."
  (Guillermo, 05:14)

• No Local Setup Required:
  A built-in VS Code editor, dependency resolution, and real-time preview are out-of-the-box, flattening the learning curve for non-developers. Quote: "You have this VS code instance, the fact that you have all your dependencies installed, the fact that this is running both with code and a preview..."
  (Claire, 08:14)

Example: Adding a Ratings System to Skills.sh (Live Demo)

• Community-Driven Features:
  Requests for verifying skill quality led to building a 5-star ratings system, with thoughtful AI prompts and safeguards against abuse. Quote: "Could we add a rating system? So let's add a five star based system rating system board skills, put it on the sidebar. Be mindful to rate-limit or prevent abuse..."
  (Guillermo, 09:07)

• Rapid, Iterative Improvement:
  The team iterates live with v0, correcting UI/UX quirks on the fly and ensuring production readiness (e.g., layout stability, data persistence).

• Production Confidence:
  PR previews are instantaneous and mirror production closely, allowing safe assessment before merging. Quote: "Now I have a production like environment...running on the production grade CDN on the production grade, rendering infrastructure, hosting infrastructure, et cetera. So now when I'm seeing that rating there, I have pretty good confidence. I was like, yeah, this is shippable."
  (Guillermo, 15:00)

Internal Usage Stats & Culture at Vercel

• Explosive v0 Adoption:
  v0 hit 3,200 PRs merged per day—100X growth in a month since early January 2026. Quote: "Starting from near zero in early January, the feature hit 3200. PR merged per day by 1-28-29, which is basically today."
  (Guillermo, 16:48)

• Fostering “Anyone Can Ship” Culture:
  Anyone in the company—marketing, design, leadership—can prototype and submit changes as PRs, with the engineering team reviewing and merging. Quote: "Until now, everyone could cook, right?...until I had this mechanism to hand it off as a pull request to the engineering team, then I was kind of like playing in La La Land."
  (Guillermo, 18:48)

• Safe Guardrails:
  Git workflows and preview environments ensure responsible shipping, even for large, high-traffic pages. Quote: "With the git workflow and opening a priority and being able to preview it, we can all build confidence that it's going to be a good change, roll it back if needed..."
  (Guillermo, 19:41)

Lightning-Round: Other AI Use Cases & Experiments

• Favorite Non-Coding AI Use Cases:

  • Image Generation: For memes, presentations, backgrounds (using internal playgrounds like Nano Banana)
  • Video Generation: Upcoming, not yet announced
  • Automated Social Cards: AI-generated OG images for product launches

• Agent to Agent Chess Game:
  Building a V0-based, AI-vs-AI chess match inspired by ESPN broadcasts, letting kids (and himself) interact, learn, and tinker with AI in a 3D visualized way. Quote: "Could V0 generate 3D code? Could it render with three JS...a live chess match? And could I have two AIs battle it out?"
  (Guillermo, 25:39)

• Family “Vibe Coding” Projects:
  Teaching his kids the real-world impact of code by connecting v0 to physical devices, like a Vesta Board, and having them see code change the real world. Quote: "I kind of taught them the concept of an API and...she was mind blown."
  (Guillermo, 29:05)

Troubleshooting and Stuck States with AI

• Debugging via AI and Collaboration:
  When AI gets stuck, Guillermo consults other models, uses built-in debug tools within v0, and sometimes “clones the repo” for a manual escape hatch. Quote: "So another thing that you can do when you get stuck is...there's a debug button...the AI itself can help itself and can give you tools to debug problems..."
  (Guillermo, 33:10)

3. Notable Quotes & Memorable Moments

• On Lowering the Barriers:
  "You're literally speaking English to the computer. So I think if you can teach them that they can express their thoughts and desires, then they can make anything happen."
  (Guillermo, 29:30)

• On Organizational Impact:
  "It reduces the friction of getting something live really, really low. The humiliation ritual of prioritization goes away and you can actually focus your time on defending the merits of an idea on the actual idea."
  (Claire, 20:37)

• On Internal Validation:
  "There's a customer zero and a customer one. Customer zero...is ourselves...But customer one is also really important, like a close design partner..."
  (Guillermo, 35:33)

4. Important Timestamps

| Timestamp | Segment Description | |-----------|-------------------------------------------------------------| | 00:00 | Friction in old product workflows; democratizing shipping | | 01:41 | Skills.sh explained; AI skills ecosystem | | 05:14 | Git-branch driven workflows directly in v0 | | 08:14 | VS Code + dependency setup in-browser; pain of local setup | | 09:07 | Live demo: adding a 5-star rating to Skills.sh via prompts | | 16:48 | v0's surge: 3,200 PRs merged/day in January 2026 | | 18:48 | Cultural shift at Vercel; anyone can ship, hand off PRs | | 22:36 | Non-coding AI use cases: image & video generation | | 25:37 | Agent-to-agent chess game; 3D code and family projects | | 29:05 | Teaching kids real-world “vibe coding” with physical devices| | 33:10 | Troubleshooting with AI, debugging, asking other models | | 35:33 | Product-market fit: dogfooding + close design partners | | 38:16 | Roadmap: v0 building agents, long-running workflows | | 40:09 | Building background/“headless” AI agents with v0 | | 42:05 | Audience Q: When will v0 support prompt-to-app-store deploy?|

5. Design Philosophy & Future of v0

• Inclusive, Productive, Safe:
  v0’s UI and branch/PR conventions are designed to make anyone—regardless of technical skill—feel like a “collaborative teammate,” but with robust engineering-grade safety and review.

• Next Steps:

  • Generative/UI rendering on-the-fly for disposable or interactive UIs.
  • Effortless deployment to mobile app stores with a single prompt.
  • Deeper agent-building support: durable workflows, retry logic, fully “headless” background agents (e.g., Slack, WhatsApp bots).
  • Scaling to more languages and real-world integrations.

6. Calls to Action

• Guillermo:
  "Get busy shipping. Try it out, give us feedback. We will fix it very, very fast because we're going to be eating our own dog food and. Yeah, share the things that you built with V0."
  (Guillermo, 34:48)

7. Audience Q&A (Condensed)

• Product-market fit validation: Customer zero (internal dogfooding) and customer one (tight design partners/community) ensure real-world utility. (35:23)
• Support for React Native and agents: The underlying “Sandbox” VMs are powerful enough to eventually support React Native and more agent types; agent-building is a key roadmap item. (36:01–39:39)
• Prompt-to-mobile-App-Store deployment: It’s a dream in the works—democratizing deployment across all platforms. (42:05)

8. Tone & Language

The episode is fast-paced, hands-on, occasionally humorous, and brimming with infectious excitement for lowering technical barriers and unlocking creativity with AI-powered workflows.

Memorable Claire-ism:
"I am in the back of my mind working on like Mama Claire’s dojo for crack little hackers." (31:32)

In Sum:
Guillermo and Claire show that truly collaborative AI—and tools like v0—are poised to erase the boundary between ideation and shipping, empowering anyone to contribute meaningfully to product development, at the speed and safety of top engineering teams.

For more:
Find full episodes, workflows, and resources at howiaipod.com